Simulated global warming affects endophytic bacterial and fungal communities of Antarctic pearlwort leaves and some bacterial isolates support plant growth at low temperatures

Abstract

Antarctica is one of the most stressful environments for plant life and the Antarctic pearlwort (Colobanthus quitensis) is adapted to the hostile conditions. Plant-associated microorganisms can contribute to plant survival in cold environments, but scarce information is available on the taxonomic structure and functional roles of C. quitensis-associated microbial communities. This study aimed at evaluating the possible impacts of climate warming on the taxonomic structure of C. quitensis endophytes and at investigating the contribution of culturable bacterial endophytes to plant growth at low temperatures. The culture-independent analysis revealed changes in the taxonomic structure of bacterial and fungal communities according to plant growth conditions, such as the collection site and the presence of open-top chambers (OTCs), which can simulate global warming. Plants grown inside OTCs showed lower microbial richness and higher relative abundances of biomarker bacterial genera (Allorhizobium-Neorhizobium-Pararhizobium-Rhizobium, Aeromicrobium, Aureimonas, Hymenobacter, Novosphingobium, Pedobacter, Pseudomonas and Sphingomonas) and fungal genera (Alternaria, Cistella, and Vishniacozyma) compared to plants collected from open areas (OA), as a possible response to global warming simulated by OTCs. Culturable psychrotolerant bacteria of C. quitensis were able to endophytically colonize tomato seedlings and promote shoot growth at low temperatures, suggesting their potential contribution to plant tolerance to cold conditions.

Figure 1

Non-metric multidimensional scaling (NMDS) endophytic bacterial (A,B) and fungal (C,D) communities of Colobanthus quitensis. NMDS plots are reported for amplicon sequence variants (ASVs) obtained from C. quitensis leaves collected in open areas (OA samples; blue) of the Antarctic site 1 (S1; square), site 2 (S2; circle), and site 3 (S3; triangle) or inside open-top chambers (OTC samples; blue) that were available in two sites (S2 and S3). Since OTCs were not available in S1, two datasets were analyzed: OA and OTC samples collected from S2 and S3 (A,C) or OA samples collected from S1, S2, and S3 (B,D). NMDS stress values are reported on each panel. Significant effects of the presence of OTCs and the collection site were found according to the permutational multivariate analysis of variance (Supplementary Table 5).

Figure 2

Heat tree of endophytic bacterial (A) and fungal (B) taxa of Colobanthus quitensis affected by the presence of open-top chambers (OTCs). Amplicon sequence variants (ASVs) affected by the presence of OTCs were identified by indicator taxon analysis with Random Forest models (Supplementary Tables 6 and 8) on C. quitensis leaves collected in open areas (OA samples) or inside open-top chambers (OTC samples) of the Antarctic site 2 (S2) and site 3 (S3). A non-parametric Wilcoxon Rank Sum test was applied to highlight differences in the relative abundances of each taxon in OTC samples compared to OA samples. Taxa with increased and decreased relative abundances in OTC and OA samples are shown respectively in red and blue, according to the colored scale legend of Log₂-transformed median proportion. For each taxon, the dimension of the node is proportional to the total number of read counts.

Figure 3

Histograms of endophytic bacterial (A) and fungal (B) taxa of Colobanthus quitensis affected by the presence of open-top chambers (OTCs). Amplicon sequence variants (ASVs) affected by the presence of OTCs and by the collection site were identified by indicator taxon analysis with Random Forest models, followed by a permutational test (Supplementary Tables 7 and 9). Relative abundances (%) of ASVs affected by the presence of OTCs were then calculated at the level of genus for samples of C. quitensis leaves collected in open areas (OA samples) or inside open-top chambers (OTC samples) of the Antarctic site 1 (S1), site 2 (S2) and site 3 (S3).

Figure 4

Growth promotion of tomato seedlings by psychrotolerant bacterial isolates of Colobanthus quitensis. The shoot length (mm, blue bars) and fresh weight (mg/plant; red bars) were analyzed for mock-inoculated plants (Mock) and plants inoculated with psychrotolerant bacterial isolates obtained from C. quitensis leaves. Assessments were carried out 15 days after seed inoculation and plant incubation at 15 ± 1 °C. Paraburkholderia phytofirmans PsJN (PsJN) was used as a positive control for its ability to promote plant growth at chilling temperatures. Mean and standard error values of 10 replicates (dishes with five plants each) from two experiments are presented for each treatment. For each parameter, colored asterisks indicate significant growth promotion in the pairwise comparisons between mock-inoculated and bacterium-inoculated plants, according to the Mann–Whitney test (P ≤ 0.05). No differences in root length were found.